You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
As grammars and Topiary code evolves separately, we are bound to break the playground at some point. We need automated tests to cover this.
Describe the solution you'd like
Either unit tests against the Wasm code, or GUI tests (e.g. Cypress) against the web playground, or both.
The text was updated successfully, but these errors were encountered:
The fact that grammars and Topiary code evolve separately is also a potential break for the CLI, and I think our tests should already catch such errors.
Do you have an example of a divergence between grammar and Topiary code that could not (should not?) be caught by the tests we already have?
That being said, having tests for the Wasm code is a good idea in general, I'm just not convinced we are threatened by divergences between grammars and Topiary code specifically, as those should be handled in the library.
The problem is when somebody updates a grammar crate (e.g. tree-sitter-nickel = { git = "https://github.com/nickel-lang/tree-sitter-nickel" }) without being aware that the Wasm grammars (e.g. web-playground/public/scripts/tree-sitter-nickel.wasm) are completely separate from this.
So if (when) a grammar crate introduces major changes, and our query files are updated to be compatible with the new grammar, the playground will stop working.
Is your feature request related to a problem? Please describe.
As grammars and Topiary code evolves separately, we are bound to break the playground at some point. We need automated tests to cover this.
Describe the solution you'd like
Either unit tests against the Wasm code, or GUI tests (e.g. Cypress) against the web playground, or both.
The text was updated successfully, but these errors were encountered: